Vivian Graubard
‘HealthCare.gov Just Accelerated It’: Vivian Graubard on the Inevitability of the US Digital Service’s Creation
| Vivian Graubard was Co-Founder and Digital Service Lead at the U.S. Digital Service for two years, beginning in September 2014. Prior, she was a Senior Advisor to the White House Chief Technology Officer. |
As a key advisor to the White House Chief Technology Officer, Vivian Graubard helped shepherd the U.S. Digital Service from an ambitious idea into a full-fledged organization. She contributed to both the high-level vision and also solving operational and tactical challenges.
During her time at USDS, Vivian worked on a range of digital transformation projects and issues, from visa modernization and gun background checks to refugee policy and crime data. Along the way she developed keen insights about the organization’s potential, but also its deficiencies. For example, the way recognition was unevenly distributed among staff and how that contributed to an “insiders vs. outside” culture.
Below, Vivian discusses friction with other federal agencies, overly simplistic narratives about USDS, and major milestones in the organization’s history.
May 27, 2019
Kathy Pham:
Viv, tell us about your journey to USDS.
Vivian Graubard:
My first job was in the Office of Presidential Correspondence (OPC), which I love very much, and other USDSers have a lifelong affinity for. I came in as an unpaid volunteer. I was 21 and had just graduated and I was trying to figure out what I was going to do next. When they saw on my resume that I had a background in tech, process engineering, and understanding user needs, Michael Hornsby and Martin Cuellar assigned me to do work that no other volunteers were doing and I was hired full-time shortly thereafter. At the time, they were redesigning their CRM, which had been custom-built and had several shortcomings. They wanted to use Salesforce, but for procurement reasons, they were forced to use a frankenstein’d custom build from a small contractor, and it didn’t really work. Every week we were filing bug complaints and requests for new features.
A year or so later, I heard that there was a job opening for the Confidential Assistant to the U.S. Chief Technology Officer, Todd Park. I interviewed for the role and I got it, and Todd was an incredible champion. Pretty quickly he elevated me out of the position of assistant and allowed me to work on other things that I cared about, which really mattered because I was also the only woman in the office. There were issues that just weren’t coming to the forefront or being naturally elevated as mission-critical work for the Office of Science and Technology Policy (OSTP), and specifically the CTO’s office.
It is important to focus on open data and energy savings, absolutely. But it’s also important to think about the role that technology and data can play in reducing violence against women and supporting vulnerable communities. It was this case of being in the right place at the right time, but also having this unique opportunity to shape the work around things that I cared about — and feeling really supported by my boss and my team.
Kathy:
How did you go from OPC to leadership in Todd’s org to the USDS? You were so pivotal. Todd had to focus on HealthCare.gov, you were basically running the office. Can you talk more about that period?
Vivian:
What a time! Of course, Todd was focusing on Healthcare.gov, but there were still other high priority portfolios across the team that needed attention. The Deputy CTO, Tom Power, was working on expanding broadband access at the time… all of that work couldn’t just stop. My role became ruthlessly prioritizing what we were giving attention to and what could wait. Healthcare.gov also failed during a government shutdown, when most people were not in the office. After the shutdown, they came back to a vastly different organization.
It doesn’t get as much attention, but the summer before HealthCare.gov failed, we ran this entire process around innovation in government. We brought together all these cabinet secretaries. Cristin Dorgelo helped plan this entire design session that was at the White House Conference Center, and Matt Collier was really instrumental in organizing it.
I mention this because USDS was going to happen one way or another, but maybe more slowly if HealthCare.gov had not crashed. We had already made a list of the top 25 services in government to redesign. We had already had sessions in the White House Conference Center and 230 ABC, and all the other rooms in the Eisenhower Executive Office Building (EEOB) where you plan these sorts of convenings. HealthCare.gov just accelerated it, sold it to the president.
Emily:
So it’s summer 2014. This thing is being built. You’ve been there for this whole prologue, and suddenly people start getting pulled in to staff this team. Do you recall when you decided, “You know what? I’m going to do this.”
Vivian:
It’s funny because so many people were working on the creation of USDS, but there was not a lot of clarity about how someone might end up actually hired by USDS.
Really, I have Erie Meyer and Brian Lefler to thank for that. It was late August or early September. Todd’s last day. No one had asked me to stay, and I’m like, “I guess I’m leaving!” September 12 officially was going to be my last day, but Brian Lefler wrote to me and said Erie had suggested we speak. He had been brought on to work on one of our first projects, which was improving the process of applying for immigration benefits. We went for a walk and Brian said, “Erie told me I should ask you about the work you’ve done.” So I told him about my work related to combating sexual assault on college campuses and other projects.
When I was done, Brian said, “I’m working with the U.S. Citizenship and Immigration Services (USCIS) and I don’t know how to get anyone to talk to me. I need you to come be on this team with me.” I told him it was my last day, and Erie and Brian then spoke with Mikey, who really supported and championed me through my time at USDS. Pretty quickly after that, the wheels were set in motion and I started working at USCIS with Brian. That really speaks to the informality of it all.
I remember thinking: “A lot of people walking around here don’t know how this building works. And the people who do know how this building works are leaving.”
Kathy:
Can you tell us about getting started, and specifically the USCIS sprint?
Vivian:
In the very early days at USCIS, it was just me and Brian trying to find areas where we could support the existing team. We would meet with the contractors and leaders who oversaw transformation efforts within USCIS. We went to the Nebraska Service Center and the Kansas City Center where millions of immigration records are stored inside caves —actual caves!
We were also working with Cecilia Muñoz’s team and the Domestic Policy Council (DPC), who were planning to launch an expanded version of DACA (Deferred Action for Childhood Arrivals) that would apply to parents of Americans (DAPA). And if we wanted DAPA to have any chance of enrolling people before the potential end of a Democratic administration, then people needed to be able to apply and be processed in a timely manner.
By then at least we had relationships, formed by stumbling around USCIS and trying to figure out what was going on, their technical infrastructure, and how it paired with policy goals. That was how we created the seedling for the sprint. We were trying to figure out: If the President announces DAPA, can we realistically build something that would allow us to enroll all of these people before President Obama’s last day in office? That gave us our sense of direction.
Kathy:
What was the structure of USDS in the beginning, and how did you see it evolve over the next few months?
Vivian:
It was very flat and very informal and very fun. Hard but fun. But I also remember feeling that I did not know what other agency teams were doing. I knew that there was a VA team, I knew that there was a healthcare team, but we were always moving so quickly and also figuring out what it was that we were doing. There was not a lot of opportunity for sharing what we were learning about what worked and didn’t work when it came to the USDS model within agencies.
So it felt like we were all in this very fun club together, but not necessarily on the same team. Almost immediately, tension with agencies would come up: Who makes the final call: USDS, OMB, another team within the White House, or the agency? Even as I say this, I’m drawing a distinction between USDS and OMB, despite the fact that USDS sits within OMB. Everyone with government experience knows there’s limited capacity, limited political capital. But the majority of USDS staff were not just new to a new agency, they were also green within government. The lack of understanding what was a real bureaucratic hurdle versus a self-inflicted one often led to frustration between people who had worked in government and people who had not.
Kathy:
If this were to be redone, there might be some level of structure or leadership that could have made things more clear. These are solvable things. Do you recall what it was like to go from a completely flat organization to the idea of Communities of Practice (COPs)? And also your role in the early days interviewing and bringing more people in?
Vivian:
I remember that moment in the basement of Jackson Place when Mikey said, “We’re about to grow really quickly and the organization is going to change radically. And you might find that this isn’t the place for you anymore because we have to add structure.” That was the beginning of the Communities of Practice (COP) and COP leads as managers. I appreciated the honesty and I liked this management style. All of the structure was necessary. But I could previously go to Mikey and tell him what I think. The prospect of not being able to do that anymore ignited a question: “Where do I belong and what does that mean for me? I’m going to have to stake a claim for myself.” I imagine that’s exactly how every startup goes. Some people feel like they’re inside and some people don’t.
Kathy:
What are you most proud of? What work stuck around?
Vivian:
I am amazed by how many people who were on the first USCIS sprint or a part of the early team stayed for years after that, and have gone back time and time again – like Dana Chisnell and so many people who I know I will fail to name. I feel really proud of the work we did at USCIS and whatever role I played in launching that team.
I worked on 15 projects during my time there. I set up the refugee admissions sprint and led the team for all of two days before being asked to pull a team together and evaluate another system. It always felt fast-paced and fun, but it was also a point of insecurity for me: What is your contribution if you were only there for the first 48 hours of a project? In a matter of four months, I worked on crime data, the refugee sprint, visa modernization, clemency, and the gun background check system.
Constantly being pulled from priority to priority taught me an important lesson, though it took me years to learn: we had this misguided belief that only one person could do certain jobs. It’s an honor to be seen in that light, certainly, but it’s also a fallacy and it contributes to burnout. The process of constantly having to find someone to whom I could pass off critical tasks — and then watching them succeed (like Raph and the Refugee Admissions Team) is a reminder that people may do the work differently and, humbly, better than you would’ve.
Emily:
You are one of the few people that saw the organization through all these different phases. What milestones stand out in your mind?
Vivian:
I love thinking about all the time before the Healthcare.gov rescue, just because there was so much possibility when thinking about what a USDS-type org might look like. That 2013 summer was exciting! Then came the fall 2013. The crash created urgency which informed the design of USDS, and who knows what it might have looked like otherwise.
The launch of 18F could be its own dissertation. I remember getting to the office one morning and someone forwarding me the press release announcing that 18F has launched, and thinking “What? We had a meeting yesterday in the EEOB about launching USDS and 18F together, and now they’ve just launched alone.” With eleven years behind us, I can see now that they were ready to go, that perhaps they wanted to be at a historically apolitical agency and not get lumped into the White House’s inherently more political orientation. But at the time, it burned! And I think it contributed to years of not understanding how we fit together, how we were different, and — most importantly — how much faster we might have moved if we worked together. Or maybe not.
Other milestones were the Roosevelt Room meeting and the State of the Union. USDS expanding really quickly, especially after President Obama’s talk at SXSW. So was the TED Talk by Haley, which had long lasting effects on how we were viewed across the White House.
Emily:
How so? What happened?
Vivian:
Initially, Todd was asked to give the TED Talk. To shine a light on someone who was in a USDS leadership position, he suggested Haley speak instead. It was very exciting at the time.
Haley’s speech was not shared with West Wing comms until the day of TED. And the speech included stats from a Veterans Affairs Powerpoint that had not been approved to share publicly. Haley was mic’d up backstage when I got a call from White House comms: “She can’t say a bunch of these things.” Imagine: You’ve rehearsed it a million times, you’re standing in the wings, and you get that call.
USDS comms made a grave error in thinking they could ask for forgiveness later. In every other scenario where a principal has to give approved remarks, it is comms’ job to keep those remarks as intact and as accurate as possible, while also being exciting and informative.
Emily:
Was there an impact on the organization as a result?
Vivian:
We were already fighting against this reputation within the White House that we thought we were the most important and critical people in that building. They felt that we were too casual, but we thought that was a symbol of breaking through red tape. The thing to keep in mind is that, when you’re a political appointee, you are united with others in your belief in the President’s vision for change. We’re all there in service of this person’s ideas for a better world that we can take part in building together. It’s not to say that other people weren’t at the White House for self-serving reasons. It’s a stepping stone in your career, absolutely. But at the end of the day, you can at least unite around this person — or that’s the expectation, at least. USDS always existed in an in-between place — not quite political appointees, not quite career staff. In theory, all there to “serve the public”, but with very different ideas of what that might mean.
Emily:
Yeah, that sentiment comes through in the other interviews.
Vivian:
It has taken me a long time to see that part of what made the work so hard was that we were all there for different reasons — but expecting that we were there for the same reason. The West Wing and decision makers for a long time had ignored the fact that we seemed self-serving and obsessed with the media, because we were doing good work. But when it became clear that we were willing to overshadow the President of the United States, that part of our comms strategy was to make this about us over a greater purpose, that was unacceptable to the West Wing.
People outside of USDS did not really buy into this idea that all of the press was necessary in order to attract top talent, because the rest of the government attracts top talent without it. It also invites this question: “If you need all this self-serving publicity to attract the right talent, then is that really the right talent?”
I’m still not sure I know the answer to that question.
Emily:
What other milestones were turning points?
Vivian:
The many times that we were shut out of agencies. We had some false starts, and people take those failures so personally that we missed out on an opportunity to understand why it wasn’t working. Every time that those things happened, there was inevitably some panic, some action that needed to be taken. There’s a lot that we could have learned about why things weren’t working, but we never felt like we had the time. There were a whole bunch of other projects that needed doing, so we’d just go focus on those things instead.
Emily:
One of the themes that stands out to me throughout these interviews is that people are focused on two different categories of goals. Category one is the product, fixing the thing. And two was the ability of government to function better. They are connected, but you can separate them.
Depending on which of those goals is your priority, you are making different decisions day in and day out. You are thinking strategically in a completely different way. The organization does have to end up being the priority, with the product as the support mechanism. That has turned into the more stated goal over time, but I have no idea what it looks like in practice.
Vivian:
Pretty early on, I came to be of the opinion that working with and through the policy councils would yield the best results. Not everyone agrees with that, but I’ve made no secret of the fact that I think working within the Executive Office of the President makes your work political, despite what people tell themselves to get through the day.
So having people who understood technology serving in roles like Senior Advisor to Susan Rice (as I did in the Biden admin) was a big deal, because it actually did shift the way the DPC was making decisions about the work we should do. That was always what I wanted us to get to, but it is very hard. It’s another reason I spent my time at New America (2016-2020) thinking through how we might apply *how* we did the work at USDS to transforming the policymaking process.
To your point, half of the organization is thinking about a tool that needs building or fixing, and the other half is thinking about that tool serving as evidence for why if the government works differently, we can do all of these amazing things.
Emily:
I have one more question related to milestones: Do you remember the agency roadshow?
Vivian:
Yes! Though I haven’t thought about it in a long time. We were going to agencies and congressional offices, sitting with them, and giving them our best pitch of what USDS was. I didn’t do many; Mikey and Haley predominantly led them.
This was, I think, another instance where we felt that pushing what felt like the buzzier Healthcare.gov narrative as a case for all-of-government transformation was the magic ticket in. One might think that the White House could just go into an agency and tell them what to do. But we know that’s not how you build lasting, meaningful change — that’s just how you tear things down.
So it’s odd to think of a White House team having to go in and “market itself” to agencies, but we did. And we quickly discovered that we did not have the agency context and were not prepared for the agency’s gotchas. Somebody like Steve VanRoekel, who had been dealing with the agency gotchas for years, would’ve been perfectly positioned for that, and he did attend a few of these. Steve had been working with the agencies for years on their tech report cards. Healthcare.gov was the loud failure, but it wasn’t the only one.
Emily:
Steve was also the federal Chief Information Officer (CIO). He has a legislated mandate.
Vivian:
Steve got us our space at Jackson Place. Steve found us the money. Steve was all in. Talk about somebody who did not get his flowers. I put Steve in the category with Laura Lynch and Matt Collier — people who actually made USDS happen, because they were coming at it from OMB and from the budget perspective and from the hiring perspective.
Still, when you think about the agency roadshow and, in general, our work with agencies, we really overestimated our authority in these spaces — that is, if you want to do things legally and equitably. When I think about the setbacks we had at USDS, it’s not a reflection of talent or the effort people were putting in, but the agency roadshow is yet another example of not being honest with ourselves about how we appeared to the rest of government. We were not setting people up to be successful when they went into these rooms.
USDSers who were new to government were coming in at a high level, seemingly of out of nowhere (to our agency partners), and with really no understanding of how government worked. Another thing we didn’t consider is that when you’re recruiting from “top tech companies,” many designers, researchers, engineers, and product managers spend their time working with peers in the same job, not with people in totally different roles, with different backgrounds, let anyone in the entirely different world of the public sector.
Emily:
That’s a really great observation. Before we wrap, why do you think telling the story of USDS has been so difficult?
Vivian:
The problem in telling the story of USDS is the myth that there is just one story to USDS. That there were five people who understood our government needed to be modernized. With USDS, there were people in the right room and the right positions at exactly the right time with the support of a president — but that doesn’t mean that our idea was novel or that we were the first people to have it. The idea that we needed to innovate in government and upgrade the way that we work with the public was not ours alone.
I am proud of my time at USDS, but the creation of USDS in some form was inevitable.